1 Comment

Summary:

Many people are familiar with ads that follow them across different websites. Now, one company says it can have the same ads show up on an individual’s smartphone too. Marketers like the idea but does it actually work? And will regulators tolerate it?

While it’s easy for advertisers to track people across different websites, the trail goes cold when a person turns off the computer and resumes surfing on a phone or tablet instead. That’s starting to change, however, as brands deploy technology that matches identities across devices.

One example is New York ad company Tapad, which claims its “Device Graph” uses 2 billion data points to find people on whatever screen they’re using. This means, for example, that a shoe company can identify a potential customer and hit them with a succession of ads — one on their work computer, another on their phone as they walk home and yet another as they look at a tablet on their couch.

In a recent interview, CEO Are Traasdahl said about 100 clients are using Tapad’s technology, including Dell and major electronics and finance firms. The appeal, he says, is that ad campaigns are more efficient if marketers know that an ad viewer on a computer and a smartphone is the same person.

Tapad’s technology may provide a boon to marketers but it also raises some obvious questions. First, how do we know it works? Mobile marketing is tricky in the first place because smartphones and tablets generate fewer cookies (the bits of computer code that indicate you’ve visited a given website); for example, Apple limits cookie collection by its Safari iPhone browser and by third party apps. This makes it difficult for brands to use so-called “retargeting” (showing an ad to someone based on what they done in the past) on a mobile device; identifying that same person across multiple devices would seem even more difficult.

According to Traasdahl, Tapad uses a sliding scale to guess whether a computer and phone user is the same person. Like all advertising, he says, there is an inherent degree of uncertainty. But he adds that, for brands, the proof is the pudding — they can look at whether multidevice targeting produced a lift in response or sales.

There is also the question of just how Tapad is tracking people. Like other companies in the mobile marketing space, such as Google Ventures-backed Adelphic, Tapad is tight-lipped about its techniques. Traasdahl did, however, say the company uses data sources like publishers’ log-in information, Wi-Fi locations and zip codes as some of the sources for its billions of data points. This is an example of what a recent MIT Technology Review article described as “reverse-engineering” our online identities.

Finally, in addition to the technology dimensions, there is the creepy factor. While companies like Tapad may provide more efficient advertising, some are going to bristle at its efforts to track them wherever they go. Tapad, like others in the advertising industry, protects “PII” — personally identifiable information — which means the “identity” that marketers see is just a random number, not your name or address. Still, there is growing concern in the media about tracking (see this week’s “Technology Turns to Tracking People Offline” in the New York Times) and in Congress where there are regular mutterings about Do Not Track legislation.

For now, the likes of Tapad and its investors (who include FirstMark Capital and Avalon) are counting on a light regulatory hand as they fine tune the online marketing machine.

(Image by Alexey Fursov via Shutterstock)

  1. I have to confess that since I was trained in economics the pursuit of consumer outliers at times appears to be driven by a mountain-climbing style desire to conquer the summit than reason or ROI. On a completely cynical level, its like finding your addicts so you can abuse them (shoe purchasing/fetish/hoarders, for example). The old standard of a message/ad needing to be seen 27 times before it reaches full awareness is still a working measure but juicing it to the point of stalking isn’t likely to produce the desired results.

    Share

Comments have been disabled for this post